1. INTRODUCTION

1.1 Terms of Reference of the Evaluation

15. The evaluation took place some six years after the initiation of country level work under the Special Programme for Food Security (SPFS). It was undertaken both in response to the request of the Governing Bodies and to meet internal management needs, and was designed with two aims, namely to:

1. Provide a credible accountability report on the SPFS, containing in-depth analysis and assessment of its continuing relevance, effectiveness in achieving results and overall cost-effectiveness; and
2. Consolidate and enhance the knowledge base of the SPFS for the future by learning from the experience to date, especially by identifying emerging issues, strengths and weaknesses.

16. The Terms of Reference (see Annex 1) covered an overall assessment of the SPFS, including South-South cooperation, and:

• Review of the efficiency and adequacy of the programme implementation process;
• Analysis and assessment of overall achievements and results;
• Assessment of the viability, sustainability and replicability of the results in the participating countries; and
• Recommendations for programme improvement.

1.2 Evaluation Arrangements and Modalities

17. A representative team of nine senior external consultants undertook the evaluation. The FAO Evaluation Service provided operational support. The Evaluation Team was composed as follows: D. S. C. Spencer (Team Leader, Sierra Leone), P. Spitz (Team Leader, Asia and West Africa missions, France), F. Anderson (Australia), M. Contijoch (Mexico), A. Maziliauskas, (Lithuania), D. Norman (USA), M. Sala (Ms) (Finland), V. S. Vyas (India), and M. Zaroug (Sudan). The team covered a wide range of expertise including agricultural economics, agronomy, animal science, farming systems, gender analysis, irrigation and water management, participatory approaches, evaluation, and rural development. Summary curriculum vitae of the team members are provided as Annex 2.

18. The Evaluation Team visited FAO Regional offices and 12 SPFS countries from each of the developing regions (i.e. Bangladesh, Bolivia, Cambodia, China, Ecuador, Eritrea, Haiti, Mauritania, Niger, Senegal, Tanzania and Zambia). Approximately one week was spent in each country by groups, which normally consisted of four consultants. In each country, the team held discussions with government, donors and FAO staff, and visited a sample of project sites using a checklist of points based on the terms of reference to facilitate their enquiries with farmers, and developmental and SPFS staff. Six members ¹ of the team convened at FAO Headquarters for two separate periods of report writing and consultations with FAO staff.

19. Countries for visits were selected by the Evaluation Team from a short-list prepared by FAO of 18 countries representing each of the developing regions. In the short-listed countries, work had been ongoing in the field for at least three years, with at least three of the four components of the SPFS. The criteria for selection ensured that the team would be evaluating on the basis of substantial experience in implementing the SPFS. There was sufficient time for some results to have been achieved, thus enabling the Evaluation Team to form a view of the prospects for impact. However, this also meant that the team only made field visits to programmes that may not have benefited from the changes that have been reflected in more recent projects whose designs may have incorporated the experience from the countries visited. Furthermore, the difficulties of the countries that, for one reason or another, had only been able to implement less than three components could not be observed first hand. To compensate for these shortcomings the Evaluation Team extensively consulted available documentation and FAO staff at country, regional office and headquarters levels. To assist the team in its work a general briefing note and individual country briefing notes were prepared by FAO. Consequently, notwithstanding the above caveat, the team is confident that it has been able to obtain a fair and representative sample of the SPFS as it has operated from the onset to the present and takes full responsibility for this independent report.

20. However, the Evaluation Team notes that it spent a relatively short time in each country and documentation on financial returns to producers as well as on the rate of adoption of technologies by non-demonstration households was inadequate. This has meant that assessment has been largely based on extensive discussions and on questioning at field level in sample sites.

1.3 Outline of the Report

21. Chapter 2 of this report provides some background on the SPFS, covering the rationale for a focus on food security and the development of the programme concept. Chapter 3 deals with the planning and design of specific SPFS programmes and the actual process of project formulation. Chapter 4 is focussed on SPFS implementation and management and considers the role of FAO, the organization and management structures, the effectiveness of national inputs, South-South Cooperation and the roles played by other international agencies and donors. Chapter 5 is devoted to assessing the pilot field operations in terms of selection of sites, target beneficiaries and technologies for testing. The approach used in implementing the SPFS field initiatives is also discussed, as are the results achieved. Chapter 6 considers the impact of the SPFS on national policies and the donor community, and also briefly deals with the cost effectiveness of SPFS initiatives. Finally, Chapter 7 brings together the material presented in the earlier chapters by summarising the background and strengths of the SPFS as viewed by the Evaluation Team. This provides the foundation, which the Evaluation Team uses to propose the approach to be applied in planning and implementing the SPFS in the future.

2. BACKGROUND TO THE SPFS

2.1 Focus on Food Security

22. At the World Food Summit in 1996, FAO member countries committed themselves to reduce the number of malnourished people by half (i.e. to reduce the number to around 400 million). The FAO Strategic Framework, approved at the 1999 Session of the Conference, reaffirmed FAO's role in assisting them to achieve this goal.

23. Malnutrition is present in all countries but the proportion in 1996-98 was 18 percent for the developing world as a whole,² whereas it was: 34 percent in Sub-Saharan Africa (rising to 42 percent in East and Southern Africa and 50 percent in Central Africa); 31 percent in the Caribbean; and 23 percent in South Asia.

24. There is a high level of coincidence in low-income countries between national food deficits and the proportion of undernourished but there are exceptions. For example, Uganda and Vietnam are not food deficit countries and have 30 and 22 percent undernourished respectively. The total food supply available in a country is the single most important determining factor in the proportion of malnourished. In general, the link between GNP per capita and food security is stronger than the link to whether a country is a net food importer.

2.2 Development of the Programme Concept

2.2.1 Programme Objectives

25. The Special Programme for Food Security, called at that time the `Special Programme for Food Production in Support of Food Security in Low-Income Food-Deficit Countries (LIFDCs),³ was endorsed by the FAO Council at its 106^th Session in 1994. In the 2000-2001 Programme of Work and Budget, the objectives for the SPFS were stated as: "assisting the Low-Income Food-Deficit Countries to increase food production and productivity on a sustainable basis; reduce year-to-year variability of production; and improve access to food, and by implication: increasing net incomes of small farmers, generation of rural employment, and reduction of poverty with due consideration to social equity and gender sensitivity."

26. All FAO member countries meeting the criteria of an LIFDC are considered eligible for assistance under the SPFS. Eighty FAO member countries currently meet these criteria. It has also been agreed that the SPFS may be extended to non-LIFDCs with alternative sources of funding from the Regular Programme. Table 1 presents data for FAO member countries with LIFDC status, as well as for non-LIFDCs which have applied for or initiated SPFS programmes.

27. At the launch of the SPFS, the reasons for chronic as distinct from emergency food insecurity and malnutrition were analysed as including four closely related factors, which are both a cause and an effect of poverty:

• Low productivity in agriculture, compounded by policy, institutional and technological constraints;
• High seasonal and year-to-year variability in food and other agricultural production which is often linked to insufficient water or inadequate water control for crop and livestock production;
• Scarcity of on and off-farm employment opportunities; and
• Inadequate and uncertain incomes in both rural and urban areas.

28. While in many low-income countries food production had not been rising as fast as population increases, food imports often did not provide a viable response, as foreign exchange was simply unavailable. Similarly non-emergency food aid was on the decline. FAO has argued that the only solution available to countries was therefore to increase food production. The SPFS was thus focused initially on staple foods and it was stressed from the outset that crop expansion would be country specific and market oriented. Emphasis was placed on high potential areas and emerging farmers who could assist in overcoming national food deficits. In 1996, emphasis was increased on horticultural products, tree crops, small livestock and fish, which could often provide the greatest opportunities for employment and income generation. They could also provide the possibility for higher value import substitution and savings in foreign exchange. The programme concept has increasingly emphasised work in lower potential areas with food deficit households and is now intended to demonstrate approaches in all agro-ecological zones.

29. The design of the SPFS, as stated by FAO, emphasised four constraints:

1. Shortage of available moisture. Except in the more humid and temperate areas of the world, soil moisture availability is the major factor affecting crop and livestock performance, limiting yields, and contributing to inter-seasonal variability in outputs, thus rendering technologies, which offer the prospect of raising productivity, uneconomic;
2. Lack of access to improved technologies. Once moisture constraints are overcome, varietal response capacity, soil fertility limitations, incidence of pests and diseases (i.e. both during production and storage) and lack of management capabilities and skills, emerge as major limiting factors;
3. Excessive dependence on a narrow range of products. The lack of diversification within many production systems exposes farmers unduly to climatic vagaries, pests and diseases and price fluctuations; leads to major fluctuations in income levels and food availability during the year; contributes to environmental degradation; may fail to take advantage of complementarities (e.g. between livestock and crops); and has negative effects on diet and health; and
4. Lack of incentives for raising output and improving sustainability. This is possibly due to a negative policy, institutional and socio-economic environment, which may combine to actively discourage farmers from expanding output (i.e. including weaknesses in agricultural services, insecurity of land tenure, price instability and an absence of, or breakdown in, rural infrastructure).

30. As originally conceived the SPFS had two phases, Phase I, the pilot phase and Phase II, the expansion phase. Phase I addressed the first three constraints listed above. Phase II was intended to give particular attention to policy and investment, addressing the fourth area of constraint. No country has, as yet, formally entered Phase II and emphasis has shifted to spreading the pilot programme to a wider range of agro-ecological zones (i.e. Expanded Phase I). It has also now been stated that policy issues should be addressed as and when they arise.

31. The SPFS apparently recognised the very substantial gap between the production levels achieved in on-farm demonstrations and trials and those of the average farmer. Demonstrations were taken as the point of entry. These were intended as a means not only of providing farmers and entrepreneurs with an insight into the possibilities but also to explore with the farmers in an ongoing participatory process, the constraints to intensification of technology and improved management systems and how these might be overcome. During Phase I the SPFS has had four standard components:

1. Development of water management and irrigation potential;
2. Intensification of sustainable plant production systems in both rainfed and irrigated areas;
3. Diversification (i.e. small animal production including poultry and small ruminants for alternative income earning); and
4. Constraints Analysis (CA).

32. South-South Cooperation (SSC) began in 1996 and from 1998, separate provision was made in the SPFS budget. It has become an important instrument in the SPFS and is designed for more advanced developing countries to send field technicians and experts to partner countries for two to three years, during which they work directly with the rural communities involved in the programme. The number of experts required is determined on a case-by-case basis, but with the objective of attaining a critical mass.

33. National management and integration into national programmes were regarded as key aspects of the programme from the point of view of national ownership, sustainability and cost-effectiveness. The programme has not thus used long-term international expertise, except that provided on a SSC basis.

34. The SPFS is currently under implementation in 62 countries of which 58 are LIFDCs. It is under formulation or awaiting start-up in a further 14 LIFDCs. A summary of the status of SPFS country programmes is provided in Annex 3.

2.2.2 Resources for the SPFS

35. Resources for SPFS activities come from countries' national resources, FAO's regular budget, various donors (i.e. principally the bilateral agencies) and International Financing Institutions (IFIs) such as the World Bank, as well as through the provision of personnel under SSC arrangements. FAO's budget includes US$ 5 million per annum of seed money for the SPFS, which is contributed to by all countries as part of their assessed contributions. Countries may also request funding for individual projects under FAO's Technical Cooperation Programme (TCP), which is also funded from the Regular Programme. This has been used in particular to support the irrigation component. Mention should also be made of the WIN project implemented in three SPFS countries and the TeleFood Programme, which makes grants of up to US$ 10,000 for projects by local people. Many of these are associated with the SPFS. Table 2 shows the distribution of funding for the SPFS in terms of actual expenditures from 1995-2000. Table 3 shows the committed budgets for SPFS programmes as of July 2001. These latter figures are not additive, as they apply to different periods and much of the expenditure will be in the future.

3. SPFS PLANNING AND DESIGN

3.1 Role of the Partners

3.1.1 Organizational Aspects and Role of FAO

36. Planning of SPFS projects has undergone substantial changes since the inception of the programme. Initially the programme began with a Coordinating Unit in the Office of the Director-General with the Director-General (DG) himself chairing an SPFS Implementation Committee. Since the SPFS's main orientation was on technology, and particularly on technologies appropriate for irrigated agriculture, technical staff in the Organization played an important role. With the expansion of the SPFS the Coordination Unit moved to the Field Operations Division (TCO) in 1997, and a Special Programme Implementation Committee (SPIC), chaired by the ADG Technical Cooperation Department, replaced the original committee, but has not met recently. The intention was for the SPFS to become a focal point for FAO's field programmes in LIFDCs. With enactment of the policy to entrust design of the technical cooperation projects, as well as of investment projects, to the FAO Investment Centre (TCI), this became the primary unit responsible for designing SPFS projects. An additional reason, and justification, for involving TCI from the very beginning was that the expected outcome of pilot projects would be the development of a country investment strategy in which TCI could play a significant facilitative role. In addition, it provided the potential for TCI to build on the knowledge it had gained through such missions in designing other projects in the same countries. The Evaluation Team was informed that initially TCI felt somewhat constrained by the rigid nature of the initial SPFS programme, although this has become more flexible over time (see Section 7.1).

37. In FAO documentation a process approach is emphasised in designing and finalising projects. This process is expected to involve not only the national government but also other stakeholders, particularly farmers. Once an agreement is reached between the potential host country and FAO, the planning process at the TCI level starts with the organization of a Core Team, consisting of relevant FAO staff from different services. The Core Team is to assist the National Programme Formulation Committee in designing the project. TCI's role tends to end once the project is designed. As there is no effective system of evaluation and monitoring for the SPFS (see Section 4.1), it does not get the necessary feedback to help it to learn and improve the design of subsequent projects.

38. With the move towards decentralisation, FAO Regional and Country Offices could play a more important role. At present these offices provide technical support at the implementation stage but generally have little say in the design of projects, although there are exceptions (e.g. in the case of Ecuador).

39. The Evaluation Team commends the participatory philosophy of SPFS and its insistence on stakeholder participation in programme design. Therefore, FAO must avoid intervening in ways that would negate the benefits of the approach. The Evaluation Team was made aware of an example of such top-down intervention in Senegal. Having allowed Farmers' Associations to participate fully in the identification of priority projects through a long consultative process, FAO intervened in a top-down manner in the approval process. Without very convincing arguments as far as the Evaluation Team is concerned, it rejected proposals that came from the grassroots (e.g. threshers and mills aimed at alleviating women's workload) and pressed for inclusion instead of problematic micro-gardens and metal silos, both introductions coming from experience in Latin America. It would appear that it was only after Farmers' Associations protested that a few mills and threshers were included in the programme. While there might be justification for inclusion of the technologies proposed by FAO, the process adopted is counter to the Organization's own claim for full participation and partnership with stakeholders.

3.1.2 Role of the Country

40. The process of project formulation outlined in the Guidelines for the Formulation of the Phase I of the SPFS gives very high priority to the national ownership of the projects. It clearly states that the responsibility for project formulation is national (i.e. the National Programme Formulation Team) with FAO playing primarily a catalytic role. Among the countries visited by the Evaluation Team (i.e. the case study countries) only a few appeared to have had a major input at the project formulation stage (e.g. China, Ecuador and Senegal). There were even examples where the formulation document was finalised after SPFS implementation activities had commenced (e.g. Bangladesh and Zambia).

41. An important pre-condition for an effective SPFS initiative is congruence between a national policy on food security and SPFS goals. Where a national policy already existed not only have the countries taken greater initiative at the formulation stage, but also they have had a greater influence in defining and elaborating the components of the SPFS and in selecting appropriate technologies (e.g., China). Also in at least one of the case study countries (Ecuador), the issue of food security was conceptualised within the national setting, enabling the SPFS to be designed in a consensual manner and in a way that was congruent with the national food security policy.

42. However, in many of the countries visited by the Evaluation Team, the role of the national governments in the planning and design of the SFPS has been rather passive. This could have been due to one or more of the following three factors:

1. Indifference on the part of the country concerning the goal of food security;
2. Initially FAO being inflexible in terms of the design of the SPFS, which inhibited open discussion among stakeholders on the structure and priorities of the SPFS during the formulation stage; and
3. The country's weak institutional and personnel base, which necessitated FAO taking a lead role in designing the project.

43. An important consequence of the country not being fully engaged at the design stage is likely to be a reduced sense of ownership and responsibility for the SPFS as far as the government is concerned. Instead the SPFS is likely to be regarded as an FAO project, and a minor one at that, because of the small level of financial resources involved. Rather than designing the programme per se, the role of FAO should be to support institutional processes such as:

• Helping in the establishment of the National Programme Formulation Team;
• Providing experts if needed;
• Assisting in capacity building in planning and project formulation; and
• Funding the preparatory tasks.

Where the above role has been adhered to by FAO, the results have been very satisfactory (e.g. Ecuador).

3.1.3 The Role of Other Donors

44. With the exception of Senegal, other international agencies have played no part at the initial design stage. In Senegal the Italian Government participated in the design of an SPFS project that it later financed (US$ 1.8m over a three years - 1997-99). It also provided financial support at the initiation of SPFS activities in Eritrea. In general, components of SPFS were `owned' or integrated by other donors at a later stage. Further discussion of the contribution of donors is in Section 4.5.

3.2 The Process of Project Formulation

3.2.1 Procedure

45. The Guidelines for the Formulation of the Phase I of the SPFS emphasise two fundamental principles underlying the formulation of SPFS projects: national ownership and stakeholder participation. A typical SPFS Preparatory Plan implements the following steps:

• Establishment of FAO Core Team made up of FAO staff from Headquarters and the appropriate Regional Office;
• Exploratory mission by the FAO mission leader;
• Review and approval of the Mission Report by the Government;
• Establishment of the National Steering Committee and National SPFS Team; and
• Formulation of the National Programme Document and Plan of Operation by the National Programme Formulation Team.

46. Generally, these steps have been taken as a matter of course. However, two other stipulations of the Guidelines were not always strictly adhered to, namely: as indicated earlier, collaboration with national partners was insufficiently emphasised; and there was insufficient consideration of the local capacity for implementing SPFS initiatives. Such deficiencies in design obviously become apparent once the implementation has started. Project designs, in general, were often found to be deficient with respect to the following:

• Did not provide a realistic assessment of institutional capabilities to implement/support SPFS initiatives within the country;
• Did not take sufficient cognisance of the expertise and experiences of other analogous development projects in the country;
• Generally did not assess the economic rationale for the proposed activities;
• Did not recognise the critical role of monitoring and evaluation; and
• Did not indicate the mechanisms for ensuring cooperation/collaboration with other institutions (e.g. research and training institutions and NGOs).

47. In the case study countries, problems relating to one or more of the above were very obvious to the Evaluation Team. Failure to adequately assess institutional capability means that the projects tend to become over optimistic and/or the same model is applied to countries and regions at different stages of institutional development. In the view of the Evaluation Team, the first step in project formulation should be an appraisal of the human and institutional capacity. A critical review of analogous national and donor-funded projects would also help in improving the design of the SPFS and the potential pay off from its implementation. Furthermore, the credibility of the SPFS would be enhanced if the potential economic and social costs and benefits were systematically assessed. As indicated elsewhere in the report, the Evaluation Team believes a lack of systematic economic analysis is a major weakness at all stages of SPFS projects. The same applies to the monitoring and evaluation component in the design of the projects, which is also considered elsewhere in the report. Finally, mechanisms for collaborating with other national organizations (training and research institutions, Farmers' Organizations, NGOs, etc.) that are important in supporting/enhancing the impact of potential SPFS activities are rarely explored and delineated in SPFS formulation exercises. Two examples of exceptions are the mention of NGOs for social mobilisation in Bangladesh, and the case of Farmers' Organizations in Senegal.

48. Although the approach advocated for designing SPFS projects is a `process' method, it has often been adhered to rigidly with little attention to innovation, or even adaptation based on experience. If, in the design of the projects, a dynamic participatory logframe approach was adopted, involving stakeholders in a participatory manner in defining the objectives, identifying the opportunities and constraints and possible solutions, and suggesting indicators to measure key variables, some of the possible pitfalls could have been avoided.

3.2.2 Programme Components

49. In a typical SPFS project, there are four major components: water management/irrigation, crop intensification, diversification, and constraints analysis. Initially, irrigation and crop intensification were the key components. Given the philosophy of the SPFS, it was logical to concentrate on water. Proper use of water results in higher yields and enables multiple cropping, both of which can be important in augmenting food grain production. Water is a critical element in sustainable agricultural development in most parts of the world. Emphasis on water, not only in SPFS projects but also in an overall agricultural development strategy, is therefore generally fully justified. Therefore, it would have been reasonable to have expected from the outset, that the SPFS would have placed greater emphasis on a wider set of water-related activities (e.g. moisture conservation, water harvesting and watershed development) rather than focussing mainly on irrigation. Even in terms of the irrigation component, it is important to recognise that farm-level irrigation (i.e. the focus of SPFS activities) is heavily influenced by the management of the larger system (i.e. above the farm level) and that the latter may turn out to be the real constraint, as is evident in the West African countries visited by the Evaluation Team. Also, in some of the new innovations introduced in farm-level irrigation, their relatively high costs in terms of infrastructure and equipment, potentially means benefits will disproportionately accrue to the farmers with more resources.

50. The activity of crop intensification associated with the irrigation component is rationalised on the basis of the need to increase production. But intensification, combined with proper water management, has additional value in augmenting land use (e.g. two or three crops per unit of land per year instead of one or two). The emphasis on scale neutral technologies, namely use of fertilisers and organic manure, and the high yielding varieties, combined with efficient water delivery systems, results in the programme being suitable for limited resource farmers, except in the labour scarce economies. The issues that relate to the strategy of intensification are at both the meso- and macro-levels. They relate to: the capacity of the research system to produce better varieties; the system of seed multiplication and distribution; the fertiliser distribution system; credit delivery; and product marketing. There are also issues relating to agricultural pricing policy, and wider environment and health-related questions. SPFS documentation stipulates that many of these issues will be deferred until at least the Expanded Phase I and usually until Phase II. However, if Phase I of the SPFS projects are bedevilled with these constraints, the prospects of moving onto Phase II are not very bright. It is indeed the view of the Evaluation Team that the issues discussed above should have been handled simultaneously with the demonstrations and other activities of Phase I.

51. Diversification as a component of the SPFS came a little later. This may be because the idea of household food security as an issue in itself was initially insufficiently appreciated (i.e. the assumption being that attaining national food security would itself ensure household food security). For resource-poor farmers, especially the landless labourers and marginal farmers, diversification of activities is essential for ensuring an adequate livelihood. In the original design the diversification component was operationalised in terms of animal husbandry, with specific emphasis on small-animal raising, activities. Fortunately, a more flexible approach was adopted at the field level in most of the case study countries. Examples of vegetable gardening, horticulture, small-pond fisheries (i.e. aquaculture), rice-fish combinations, poultry etc., are as numerous as pig or goat enterprises. These activities directly address the income earning capacity of food-insecure households, provided they are specifically focussed on them. As women operate many of these small enterprises, this component, more than any other activity under SPFS, helps in empowering women.

52. Constraints Analysis (CA) is the fourth principal component of the SPFS. In this discussion on the design of SPFS activities it is important to emphasise that in the original documentation of the SPFS it was not envisioned that it had any role to play in the design stage. Logically the Evaluation Team believes CA should precede the launch of the project (i.e. be used in the design of the SPFS activities), as well as being repeated at different stages to ensure mid-course corrections. Because, as a rule, at least in the case study countries, it has not been done before the start of SPFS activities, the design of the SPFS has not benefited from CA. However, some benefits from CA remain if, as a result of such exercises during implementation of SPFS, constructive and useful modifications are made in subsequent activities. Assertions made in some quarters that CA will be particularly relevant in Phase II of SPFS when policy issues will be resolved, misses the point that many constraints faced by producers are at the micro- or the meso-levels and do not necessarily involve major policy changes for their resolution. One further point is worth mentioning with reference to CA. While appreciating that CA was added as a component sometime after the SPFS was launched, the Evaluation Team is not convinced that CA should itself be viewed as a component of the SPFS. The other components all relate to tangible productive ways/means of improving food security. CA on the other hand is a methodology or, perhaps more accurately, a package of methodologies, to aid in helping to determine what should be done. Therefore, although CA is important as far as the SPFS is concerned, it should not be viewed as a component in the sense in which the term is used in the SPFS.

53. At the country level, apart from these components, several additional activities have progressively become part of the lexicon of Phase I SPFS activities. In newer projects this tendency of adding more components is increasing (see Section 7.1). Quite a few of these `satellite' activities were sponsored or supported by TeleFood, some came as additional TCP projects, and others were initiated by the national teams themselves. However, in the case study countries, these were generally initiated after the original SPFS design activity, although the Evaluation Team understands, these are now more likely to be incorporated in the design stage in countries where SPFS activities have been initiated more recently (see Section 7.1).

54. The Evaluation Team recognises that there are countries (e.g. Cambodia and Ecuador) in which different components of the SPFS were well synchronised and integrated. However, integration of the different components to generate synergy remains an issue. Conceptually, all the three `product' components are complementary, but in practice, apart from water and intensification, there was usually very little integration between the principal components, and sometimes even only a limited geographical overlap. As noted earlier in the section, introduction of the different components came at different times. For example, in an otherwise strong project, China, diversification came at a later stage and was not integrated with water management or crop intensification activities in the same villages.

3.2.3 Approach to Upscaling and Mainstreaming

55. One country (Senegal) has expanded Phase 1 activities to achieve national coverage. In the Expanded Phase I, components are implemented as appropriate in each agro-ecological zone, with only one or two components being implemented in the majority of sites. To the Evaluation Team this indicates a welcome degree of flexibility in the design of the Expanded Phase I. However, in most case study countries a major lacuna in the SPFS design is the absence of a strategy for `upscaling' the project. A pilot project has a raison d'être in its potential for expansion or replication. The Evaluation Team is aware that, in a number of countries, proposals for expanding the number of sites are being considered. But such proposals are often in the nature of `more of the same', in a few more locations, which therefore are not likely to address the diversity of production (i.e. biophysical and socio-economic) environments in the countries.

56. One of the primary objectives of the pilot projects should be `mainstreaming' the approach and the content of the SPFS for which a multi-pronged approach is needed. Five groups of actors are important in this effort:

1. An initial condition for this type of approach is already available in many countries where the projects are located in the Directorate of Extension. To the degree that extension personnel are convinced of the SPFS approach, there are possibilities of replication subject to the appropriate adjustments as implied in the preceding paragraph;
2. Another major group that require convincing about `mainstreaming' the SPFS are the policy makers. The FAO representative (FAOR) and staff in the FAO Regional Offices can be important in this effort but it is not clear as to whether this role has been explicitly assigned to them;
3. The third group to be sensitised is the donors. As was reported earlier, with a few exceptions, the donor community has tended to be largely indifferent;
4. In most societies, particularly in open societies, the opinion leaders -- journalists, intellectuals, professionals -- matter. They need to be convinced about the relevance and efficacy of the projects; and
5. However, the most important group are the project beneficiaries. If they have benefited from the SPFS approach, they will help in spreading the message.

57. If the SPFS is considered not only an important programme but also a distinct and superior approach, the strategy to `spread the word' should be explicitly addressed. This, of course, assumes that lacunae in design and implementation are taken care of, and a sound project results that is worthy of `upscaling' and `mainstreaming'.

____________________________

1 Messrs Spencer, Maziliauskas, Norman, Spitz and Ms Sala took part in two missions each, while Mr Vyas participated in one mission, and all six were involved in overall report writing. Messrs. Anderson, Zaroug and Contijoch took part in one mission each but were not present for report writing.

2 All figures for undernourishment refer to the State of Food Insecurity in the World (SOFI), 2001, FAO, Rome.

3 Defined as those countries with a negative trade balance in cereals and per capita incomes of a level, which qualifies them for World Bank lending on IDA terms (i.e. countries which in 1999 had per capita GNP of US$ 1,445 or less).